Loan phonology (review)

For a number of years, loan phonology has been viewed as fertile ground for investigating the nature of phonological grammars. Loan phonology is a most useful collection of articles on the subject with a very clear introduction by the editors. The basic relevance of borrowings for phonology, and more generally for cognition, stems from classical work about the phoneme by linguists like Leonard Bloomfield, Roman Jakobson, Edward Sapir, and N. S. Trubetzkoy. It builds on the idea that the phonological system of a language acts like a 'sieve' for speakers of a language, shaping their perception and consequently their production not only of sounds produced within their own language but also of the sounds of other languages, and even leading in one interpretation to what has been called phonological DEAFNESS (see Polivanov 1931 for an early statement of the issues). In the wake of The sound pattern of English by Chomsky and Halle (1968), there has emerged an active tradition of work that uses borrowings to test hypotheses on the nature of phonological representations and processes (e.g. Hyman 1970, Lovins 1973). In recent years, the challenge from phonetic approaches to phonological issues has given new impetus to the topic, comprehensively covered in this collection of essays.

One position dubbed the PHONOLOGICAL STANCE is discussed by CAROLE PARADIS and ANTOINE TREMBLAY in 'Nondistinctive features in loanword adaptation: The unimportance of English aspiration in Mandarin Chinese phoneme categorization', which builds on a great deal of earlier work going back to the late 1980s (e.g. LaCharité & Paradis 2005). The authors assume that bilingual speakers play a crucial role in the borrowing process whatever their motivation for adopting foreign words may be (e.g. filling in a lexical gap, acquiring prestige, indulging in wordplay). For instance, a speaker S of the recipient language (L1) borrows a word from the donor language (L2). The idea that Paradis and Tremblay advance is that S retrieves the underlying form of the borrowed word from her/his L2 long-term memory and grinds it through the grammatical rules or constraints of the L1 system to generate a surface form. In doing so, a number of repair strategies will be used to allow the surface form to conform to L1 templates: the borrowed word will be nativized according to the grammatical system of L1, its PHONOLOGY. On the basis of a well-known, extensive database of loanwords constituted at Laval University, it is argued that phonetic details are not central to the adaptation of loanwords, which is essentially driven by phonological constraints. Using a corpus of 500 stops included in 371 borrowings from English into Mandarin Chinese, Paradis and Tremblay show that stop aspiration usually considered allophonic in English has no influence on phoneme categorization in Mandarin Chinese, despite the fact that the latter makes a phonemic distinction between aspirated and nonaspirated stops. Mandarin Chinese speakers should be predisposed to distinguish aspirated from unaspirated stops (e.g. pit [p^h It] from spit [spIt]); but what they do is map all English voiceless stops onto aspirated stops, while English voiced stops yield voiceless unaspirated stops in Mandarin Chinese.

An opposite position is defended by HYUNSOON KIM in 'Korean adaptation of English affricates and fricatives in a feature-driven model of loanword adaptation', which takes into account that L1 speakers have variable competence in L2, including the possibility that they have poor or no knowledge of L2. The adjustments and adaptations that can be observed in loanword adaptation take place during perception and learning, a scenario that is referred to in this volume as nativization-through-perception or the perceptual stance model. The perceptual stance model is itself compatible with two extreme views. At one extreme, it can be assumed that the perception by S of the L2 is driven by the phonological grammar of the L1 and determines the adaptation of L1 borrowings in L2. Kim defends this view and argues that the L1 grammar...